Parisar-Netra - Net-Zero Campus Intelligence System
🏆 1st Prize (₹25,000) at CITI-ZEN 2026 TERRABYTE Hackathon. An AI-driven cyber-physical platform using Edge IoT and Google Gemini to autonomously eliminate 'Zombie Loads' in university campuses.
Overview
Parisar-Netra is an autonomous Net-Zero Campus Intelligence System built by Team Detox at the CITI-ZEN 2026 TERRABYTE Hackathon held at Alliance University, Bengaluru. The project won 1st Prize (₹25,000).
The system transforms university campuses into "Smart Energy Cells" by combining Edge IoT hardware with AI-powered decision making to autonomously detect and eliminate wasteful energy consumption — known as Zombie Loads.
Key Impact Metrics
- ⚡ Up to 15% Energy Reduction across monitored campus zones
- ⏱️ < 2 Month ROI from energy savings alone
- 🌍 Offsets 0.82 kg CO2 per kWh saved, advancing Bengaluru's Net-Zero climate goals
Problem Statement
University campuses waste significant energy through unoccupied classrooms with running HVAC, lights left on in empty hallways, and equipment drawing power overnight. These "Zombie Loads" account for up to 30% of total campus energy usage.
Solution Architecture
Edge IoT Layer
- ESP32 microcontrollers with environmental sensors (temperature, humidity, light, occupancy)
- Local edge processing for real-time decisions without cloud dependency
- Privacy-preserving occupancy detection using on-device TensorFlow.js vision models
AI Intelligence Layer
- Google Gemini 1.5 Flash for multimodal energy auditing — analyzing sensor data, camera feeds, and weather forecasts simultaneously
- Predictive HVAC optimization based on class schedules and occupancy patterns
- Dynamic daylight harvesting to reduce artificial lighting usage
Dashboard & Control
- Real-time campus energy monitoring dashboard built with Next.js
- Automated control signals to HVAC and lighting systems
- Alert system for anomalous energy consumption patterns
Tech Stack
- Frontend: Next.js, React
- Backend: Node.js, Express
- Hardware: ESP32, Embedded C++
- AI/ML: TensorFlow.js (Edge Vision), Google Gemini 1.5 Flash API
- IoT Protocol: MQTT for sensor communication
- Database: Time-series data storage for energy analytics
Hackathon Context
The CITI-ZEN 2026 TERRABYTE Hackathon challenged teams to build solutions addressing Bengaluru's urban sustainability goals. Our approach stood out for its practical, deployable architecture — combining affordable IoT hardware with cutting-edge AI to deliver measurable energy savings with a rapid return on investment.